Back to FabImage Deep Learning website

You are here: Start » Computer Vision » Deep Learning » DL_DetectAnomalies2_WithThresholds

DL_DetectAnomalies2_WithThresholds


Header: FILDL.h
Namespace: fil
Module: DL_DA

Executes a Detect Anomalies 2 model on a single input image.

Syntax

C++
C#
 
void fil::DL_DetectAnomalies2_WithThresholds
(
	const fil::Image& inImage,
	const fil::DetectAnomalies2ModelId& inModelId,
	float inT1,
	ftl::Optional<float> inT2,
	float inScaleHeatmap,
	fil::Heatmap& outHeatmap,
	bool& outIsValid,
	float& outScore,
	bool& outIsConfident,
	ftl::Conditional<fil::Region>& outRoi
)

Parameters

Name Type Range Default Description
Input value inImage const Image& Input image
Input value inModelId const DetectAnomalies2ModelId& Identifier of a Detect Anomalies 2 model
Input value inT1 float 0.0 - 100.0f 'Good' threshold value
Input value inT2 Optional<float> 0.0 - 156.0f 'Bad' threshold value, if not set 'Good' threshold will be used.
Input value inScaleHeatmap float 0.0 - 10.0 1.0f Modify visualization of the output heatmap. This does not affect outScore.
Output value outHeatmap Heatmap& Returns a heatmap indicating found anomalies
Output value outIsValid bool& Returns true if no anomalies were found
Output value outScore float& Returns score of the image
Output value outIsConfident bool& Returns false if the score is between T1 and T2
Output value outRoi Conditional<Region>&

Requirements

For input inImage only pixel formats are supported: 1⨯uint8, 3⨯uint8.

Read more about pixel formats in Image documentation.

Errors

List of possible exceptions:

Error type Description
DomainError Empty image in DL_DetectAnomalies2_WithThresholds.
DomainError Not supported inImage pixel format in DL_DetectAnomalies2_WithThresholds. Supported formats: 1xUInt8, 3xUInt8.